Extreme Vocal Dynamics: What are you doing?

[Classic]

For Country Music - Mixing ITB

I've worked with plenty of dynamic vocals, but an album I'm currently mixing is definitely challenging me. On every song, the vocals frequently bounce back and forth between super-soft whisper to full-blown singing for each line.

When I tracked the vocals, I used a little bit of compression (about 3-4 db), and now I've got two approaches going on, neither I'm really excited about:

1. Stack 3 compressors (LA2A - 1176 - LA2A) about 3db each / Then ride fader where needed, which is all over the map.

This is normally how I approach vocals, and I'm getting the sonic characteristics that I want, but it's just not fitting smoothly in the mix like most of the contemporary country vocals you hear (meaning totally squashed). So...

2. VOX SQUASH - same 3 compressors, only this time, each one screaming / Minor fader automation.

This definitely takes care of it, but sonically, I'm not really digging it. Works for voices like Radney Foster, but not in this particular case. Hard compression just wears on the ears and does something to the mids / highs I'm not a fan of. Some creative EQing takes care of most of it, but I don't know... just isn't sexy.

It's a tough call because I hear a lot of records that are squashed that sound great, but I can't help but think the entire time I'm listening, "Did they really have to kill his vocal so much that you can hear the compressor pumping??" Or even better, when the dynamics are sometimes reversed sounding - soft parts slightly louder than the loud parts. That always messes with my head.

Anyway, how do you approach extremely dynamic vocals to get a natural, smooth sound that fits in the mix?

[warhead]

I'd compress to taste tonally and with as much correction as I could accomplish, then draw the rest in Nuendo as intense automation is required. The only thing I notice about having to get down on a microscopic level, working in several second increments mixing a song is, to take regular breaks by listening to the entire song to keep the context right. You can get lost making all those little moves by not hearing it as a listener: start to finish.

War

[Mayor999]

Perhaps you should go to something other than a LA-2A?

I personally have never been a huge fan of the LA-2A set for stun sound... ( your option 2)

I love LA-2A's for tracking, but something with more control may be able to help you shape the dynamic envelope you are looking for in the mix....



You could also cut the soft parts onto another track and the loud tracks onto their own as well and treat them differently....

Some ITB solutions are handy... I really dig RVox as a "plug-in and play" plug and then dig into RComp for fine tuning attack and release.... Although you can never have too many distressors if you are looking for a OTB solution...

Sometimes though you just need to really rock a lot of volume automation.

[u b k]

itb is what it is, i never could get a vocal to behave and retain depth with the plugs.

if you go otb, you got:

atomic squeezebox

gml 8200

distressor

mpressor

525

all of those will do 15-20db of reduction without breaking a sweat. the first two are invisible; the third and forth pull the sound forward; the last one spreads it out wide.

for better or worse, this is what's true in my world: big money vocals cost big money.


gregoire
del
ubk
.

[Mike Derrick]

What kind of mic are ya using?

If it's a condensor, I'd switch it up to a dynamic (or maybe even a ribbon.)

If you'd rather keep the condenser, what about vocal mic technique by the performer, is he or she good at playing the sweet spots of the mic according to their singing dynamics? I find it's not always a great idea to have them moving, but small changes in the distance to the mic for the appropriate volume level being delivered can sometimes help tame some of the extreme dynamics.

I assume the tracking compression is outboard gear. If you like your tracking compression, why not use more compression before going into the box, especially since you already know you're likely gonna comp even more later.

I like volume automation for evening some vocal lines out for itb before going too crazy with compression. Sounds more natural to me in most cases.

[Full Clip Audio]

You can slam a Vac Rack REALLY hard and it still sounds really transparent. Even better yet would be to track with a dynamic directly below or above the condenser and capture the performance on both so you can switch between the two when appropriate. That and riding the pre when tracking should do all you need next time around.

[ThomasWho]

When I`m faced with this I always track with little or no compression , other than the logic compressor (no complaints about latency from singers yet).

At mix I automate every line for volume , going INTO hardware compressors. I make shure I`m hitting the compressor the right way for every line.

For extremely dynamic singers I haven`t found a pleasing way to to it "automatically" with comps , without automation.

Thomas

[gurubuzz]

Originally Posted by ThomasWho

When I`m faced with this I always track with little or no compression , other than the logic compressor (no complaints about latency from singers yet).

At mix I automate every line for volume , going INTO hardware compressors. I make shure I`m hitting the compressor the right way for every line.

For extremely dynamic singers I haven`t found a pleasing way to to it "automatically" with comps , without automation.

Thomas

Well said Thomas. feed the compressor a healthy diet instead of letting the compressor digest all the crap. On the loudest notes the compressor should do little and raise the quiet words instead of trying to kill the loud ones (upward Compression). Then run another compressor for colour/effect.


In the days before you could get third party plug ins, I was a master at manual compression by editing wave files, I would find the loudest peak (logic had a search peak function) and I would select at Zero crossings and turn down the peak. This could take a couple of hours for a whole vocal take..

But today with Cubase there is so much choice on gain, the quickest is to snip either side of a loud peak and and pull down the wave, and same for turning up quiet ones. Then you have the pen tool to draw an envelope then you also have good old Automation...feeding a bus comp.




BUT when Tracking try to use 2 mics one close with a lot of gain and one a little further away with less gain..

And your responsibility is also to train on microphone technique.

Teach them how to lean back on loud bits and lean forward with cheek touching mic for the whispers....

Another trick is record the "practice take" and mute it, watch the wav as you record on another track and you will be able to "look ahead" and ride the fader at the right time....

[Classic]

Originally Posted by Mayor999

Perhaps you should go to something other than a LA-2A?

Did some experimenting yesterday, and after listening, it seems that stacking an 1176 (8:1, fast attack/med release), then another at 4:1 med attack/med release, then an LA2A at about 2-3 db, then some creative automation produced the best results.

Like warhead said, easy to lose perspective on the mix when doing detail work.

Thanks for all of the replies. All great suggestions, and it's cool to hear what everyone else is doing for situations like this one.

As for what mic was used, twas a TLM49. SM7b sounded pretty good and probably would have been better for mix time (I like how it takes compression and EQ), but the TLM49 sounded obviously better.

[DesertDawg]

Originally Posted by Classic

For Country Music - Mixing ITB

I've worked with plenty of dynamic vocals, but an album I'm currently mixing is definitely challenging me. On every song, the vocals frequently bounce back and forth between super-soft whisper to full-blown singing for each line.

Why don't you just clone the track & treat the whisper/soft part completely differently than the louder parts? Mute or trim in the appropriate places.

[mixman499]

i generally put my manley elop....sometimes BOTH channels...and then often with MORE compression from my dakings....and even more times with some purple 76 love....to task for this issue.

recently, i mixed in a room with the vac rac tsl-3 (i think that is the model #) knobs cranked....it sounded AMAZING.

my pennies are being saved.

i think with one of those, i could stick my manley back on bass...and keep the dakings on drums !!

an AMAZING piece of gear for getting vocals RIGHT UP THERE.

best,

jchristopherhughes

[T_R_S]

Waves MAXX Volume

[u b k]

i'm designing a compressor that addresses this very issue. it'll do >20db of reduction on a crazy out of control voice and sound like it's doing none at all. no sibilance, virtually no artifacts. it's impossible to make it sound bad.

it has one control: more, or less.

aes 2008.


gregoire
del
ubk
.

[Musiclab]

Originally Posted by u b k

i'm designing a compressor that addresses this very issue. it'll do >20db of reduction on a crazy out of control voice and sound like it's doing none at all. no sibilance, virtually no artifacts. it's impossible to make it sound bad.

it has one control: more, or less.

aes 2008.


gregoire
del
ubk
.

Greg I like to hear that.

[Musiclab]

If your mixing on a console I'd spread the vocal across different faders, with compressor set for the soft lines and another channel with compressor set for the loud bits, this way
you can keep the compression to a level of sanity. If your mixing in the box, while I agree with UBK, you could try a similar approach

[DesertDawg]

Originally Posted by Musiclab

If your mixing on a console I'd spread the vocal across different faders, with compressor set for the soft lines and another channel with compressor set for the loud bits,

He's not (see the first line of his post... referencing ITB), which is why I recommended cloning the track & treating each part (soft vs. loud) differently, then either snip, or automate mutes on each track, as needed. Same thing as you're suggesting, of course, just the ITB version of such.